AI and WANCode

What WANCode does — and doesn't — do with AI

• WANCode doesn't call any AI service. Your code, progress and answers are never sent to an AI model by WANCode.
• Checks, hints and explanations are written by people and run in your browser — they aren't generated on the fly.
• The Prompt library assembles prompts locally: you fill in the fields and your browser builds the text. Nothing is sent anywhere until you copy it and paste it into an assistant of your choice — at that point, that service's own terms and privacy policy apply.
• If WANCode ever adds an optional AI tutor, it will be opt-in, clearly labelled, and will tell you exactly what is sent and to whom before anything leaves your device. Every lesson will keep working without it.

Using coding assistants responsibly

AI assistants can explain, suggest and draft code quickly. They can also be confidently wrong. The skill that matters is staying in charge: you decide, you verify, you understand.

• Ask for small changes. A 10-line diff you can review beats a 300-line rewrite you can't.
• Ask for explanations, assumptions and tests, not just code. Our prompts are written that way on purpose.
• Verify everything. Run it, test it, read every line. If you can't explain a line, don't ship it — ask about it.
• Check facts that go out of date — package versions, APIs, security advice — against official documentation.
• Never paste secrets: passwords, API keys, tokens, .env files, customers' personal data, or company code you aren't allowed to share.
• Keep learning the fundamentals. You can only review code in a language you can read.